JP3398928B2 - Network adapter device - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a network adapter device for a computer, and more particularly to a data transmission / reception system for an internet telephone using a network adapter device having an Ethernet interface.

[0002]

2. Description of the Related Art FIG. 15 shows the configuration of an Internet telephone using a conventional network adapter device and voice adapter device. The network adapter device 18 including the Ethernet input / output interface unit 20 is a type of computer expansion card, and is an IEEE80
Ethernet device driver 1 on the host computer that receives Ethernet frames conforming to 2.3
6 and a function of converting the data passed from the Ethernet device driver 16 on the host computer into an Ethernet frame and transmitting it to the Ethernet.

A voice adapter device 19 having a voice input / output interface unit 21 is connected to a microphone and a speaker,
It has a function of passing the voice input from the microphone to the voice device driver 17 on the host computer and a function of passing the data passed from the voice device driver 17 on the host computer to the speaker and producing a sound.

When an Internet telephone is constructed using such a conventional network adapter device and voice adapter device, the frame input from the Ethernet is no different from normal data, and therefore is passed to the Ethernet device driver 16 as it is. And is passed to the Internet telephone service application 1 through the protocol processing unit 2 corresponding to the protocol used by the Internet telephone.

In the Internet telephone service application 1, the frame is passed to the audio device driver 17. The audio device driver 17 passes the frame to the audio adapter device 19, and an audio output device such as a speaker connected to the audio adapter device 19 is rung.

On the contrary, the voice input / output interface unit 2
Input data from a voice input device such as a microphone connected to 1 is passed to the voice device driver 17, and the voice device driver 17 passes the data to the Internet telephone service application 1.

In the Internet telephone service application 1, the data is passed to the protocol processing unit 2 corresponding to the protocol used by the Internet telephone. The protocol processing unit 2 passes the data to the Ethernet device driver 16. The data is transmitted to the Ethernet through the network adapter device 18.

Another network adapter device is disclosed in Japanese Patent Publication No. 9-2767085, but this device does not have a voice input / output interface function realized by a voice adapter device. All the received data are passed to the protocol processing section through the Ethernet device driver. On the contrary, in order to transmit data on the Ethernet, it is necessary to pass the data to the adapter device through the protocol processing unit and the Ethernet device driver.

[0009]

However, in the above-mentioned conventional network adapter device, all the data received by the adapter device is passed to the protocol processing section 2 through the Ethernet device driver 16. Therefore, even when performing a voice receiving service typified by Internet radio broadcasting,
The voice data received by the adapter device is passed to the protocol processing unit 2 through the Ethernet device driver 16, and the service-specific application 1
The application 1 sends the data to the voice adapter device and outputs the voice, so that there is a problem that the delay from the voice data reception to the voice reproduction becomes large.

Therefore, even when the simultaneous reception service of audio and video represented by Internet TV broadcasting is executed, the audio data and video data received by the adapter device are processed by the protocol through the Ethernet device driver 16. It is passed to the part 2 and further to the service-specific application 1, and the application 1 sends the data to the audio adapter device or the video adapter device and outputs the audio or video. There is a problem in that the delay from time to time and the delay from video data reception to video reproduction become large.

Further, in the above-mentioned conventional network adapter device, all frames to be transmitted to the Ethernet are also passed to the Ethernet device driver 16 via the protocol processing unit 2 and further passed to the adapter device before being transmitted. To be done. Therefore, even when a voice transmission / reception service represented by an Internet telephone or the like is executed, the voice data received by the adapter device is passed to the protocol processing unit 2 through the Ethernet device driver 16, and the service-specific application 1 And the application 1 sends the data to the audio device driver 17. Further, the data is passed to the voice adapter device, and the voice is output to a voice output device such as a speaker. Conversely, voice data from a voice input device such as a microphone is passed from the voice adapter device to the voice device driver 17, and the voice device driver 17 passes the data to the service-specific application 1. The service-specific application 1 passes data to the protocol processing unit 2 and the protocol processing unit 2
Also sends data to the Ethernet device driver 16
Pass to. The Ethernet device driver 16 passes data to the network adapter device and transmits the data. Therefore, the delay from the reception of audio data from Ethernet to the audio reproduction through the audio adapter device and the audio adapter device There is a problem that the delay from the input of voice data to the transmission of voice data through the network adapter device becomes large.

The present invention solves such a conventional problem, and not only has an Ethernet adapter function, but also has a small delay from reception of audio / video data to audio / video reproduction and high-quality audio / video. It is an object to provide a network adapter device capable of providing a reception service. Another object of the present invention is to provide an internet telephone service with a small delay from voice data input from the voice interface unit to data transmission to Ethernet.

[0013]

In order to solve the above-mentioned problems, the present invention according to claim 1 is a network adapter device for connecting a computer and an Ethernet, which is connected to a computer to provide various control information and A computer interface unit for transmitting and receiving data, an Ethernet input interface unit for connecting to Ethernet and receiving Ethernet frames, an Ethernet output interface unit for transmitting Ethernet frames, and an audio output device for connecting audio data. The difference between the voice output interface unit for outputting and the preset pattern for the frame input from the Ethernet is determined, and when the patterns match, the data of only the preset offset value is removed from the beginning of the frame, remaining A frame filter section having a function of transmitting a part to the audio output interface section and transmitting the frame as it is to the computer interface section when there is no match, only an input frame from Ethernet having a preset pattern The frame filter section extracts the frame, outputs only that frame to the audio output interface, passes the other frames to the protocol software on the computer through the computer interface section, and outputs the data sent from the computer interface section to the Ethernet output frame. It is characterized in that it is directly output through the Ethernet output interface unit.

According to the above arrangement, the frame pattern information for filtering in advance by the frame filter section,
For example, if the protocol header information is set, out of the frames received from the Ethernet input interface unit, only the frame that matches the frame pattern information is extracted, and based on the offset value information for preset header deletion, After deleting the data from the head of the frame to the offset value, the data is transmitted to the audio output interface unit. Other frames are passed to the Ethernet device driver on the host computer as a normal network adapter device does. From the above, of the frames received from the Ethernet interface, for frames with a specific frame pattern, the frame header is removed on the adapter device without passing through the software on the host computer, and the raw audio data is extracted, It can be passed to the output interface unit, and the delay from frame reception to voice output can be reduced.

According to a second aspect of the present invention, in addition to the Ethernet input interface unit, the Ethernet output interface unit, the computer interface unit, the audio output interface unit, a video output device is connected, and video data is connected. Of the video output interface unit for outputting and the input frame from the Ethernet input interface unit, the difference between the preset audio pattern and the video pattern is determined, and if they match the audio pattern, Has a function to send an input frame to the audio output interface section, send an input frame to the video output interface section if it matches the video pattern, and send data to the computer interface section if neither matches. Frame filter with Of the input frame from the Ethernet, only those having a preset pattern are extracted in the frame filter section, data corresponding to the preset offset value is removed from the beginning of the frame, and the remaining portion is output to the audio output interface section or Output to the video output interface unit, other frames are passed as they are to the protocol software on the computer through the computer interface unit, and for output frames to the Ethernet, the data passed from the computer interface unit is directly output through the Ethernet output interface unit. It is characterized by According to the above configuration, if frame pattern information for filtering in advance in the frame filter unit, for example, protocol header information, is set to two types for video and audio, among the received frames from the Ethernet input interface unit, Only a frame that matches the video frame pattern information is extracted, and based on preset offset value information for header deletion, data from the beginning of the frame to the offset value is deleted, and then the video output interface unit Send to. Similarly, of the frames received from the Ethernet input interface unit, only the frame that matches the voice frame pattern information is extracted, and based on the preset offset value information for header deletion, After deleting the data up to the offset value, the data is transmitted to the audio output interface unit. Other frames are passed to the Ethernet device driver on the host computer as a normal network adapter device does. From the above, among the frames received from the Ethernet interface, for frames with a frame pattern set for specific video and audio, the frame header is set on the adapter device without going through the software on the host computer. Get rid of,
Raw video data or audio data can be taken out and passed to the video output interface unit or the audio output interface unit, respectively, and the delay from frame reception to video output or audio output can be reduced. This makes it possible to provide a high quality video reception service. In particular, the amount of video data is large, and if it is passed to the video output interface section via software on the computer as in the conventional method, the CPU on the computer
Although resources and memory resources are wasted, according to the method of the present invention, since the video data is directly passed to the video output interface section on the adapter, CPU resources on the computer,
It becomes possible to receive video data without wasting memory resources and the like.

The present invention according to claim 3 provides a filtering function based on preset pattern information for a frame received from Ethernet by the Ethernet input interface unit, the frame filter unit, and the voice output interface unit. A voice input interface section that has the function of receiving input data from the voice input device and transmitting it to the frame generation section, and adds preset header information to the data received from the voice input interface and sends it to the frame multiplexing section. A frame generation unit having a function of transmitting, a transmission frame from the computer interface and a transmission frame from the frame generation unit are received, and a frame multiplexing unit having a function of multiplexing them and transmitting them to the Ethernet output interface unit, Into a preset Ethernet frame data from the audio output device connected to the voice input interface, characterized in that that can be transmitted. According to the above configuration, the voice input interface unit receives the data from the external voice input device such as the microphone through the voice input interface, and further sends the data to the frame generation unit. The frame generation unit holds header information to be set in advance, and the frame generation unit adds the header information to the beginning of the data received from the voice input interface to form a frame that can be output by Ethernet. Hand over to the department. The frame multiplexing unit multiplexes the frame generated by the frame generating unit and the data requested to be transmitted by the Ethernet device driver on the host computer through the computer interface unit, and passes it to the Ethernet output interface unit. The Ethernet output interface unit sequentially sends the frames received from the frame multiplexing unit to the Ethernet. From the above,
Of the frames received from the Ethernet input interface section, for frames with a specific frame pattern, remove the frame header on the adapter device without going through the software on the host computer,
It is possible to take out the raw audio data and pass it to the audio output interface section, and it is possible to reduce the delay from frame reception to audio output. Also, voice data from a voice input device connected to the voice input interface unit can be directly framed and transmitted on the adapter device without going through software on the host computer. It is possible to reduce the delay until frame transmission.

According to the present invention of claim 4, a pattern preset by the Ethernet input interface unit, the frame filter unit, the audio output interface unit and the video output interface unit is set for a frame received from Ethernet. A video input interface section that has a filtering function based on information, and has the function of receiving input data from a video input device and transmitting it to the frame generation section, and adding preset header information to the data received from the video input interface A frame generation unit having a function of transmitting to the frame multiplexing unit, and a function of receiving a transmission frame from the computer interface and a transmission frame from the frame generation unit, multiplexing them, and transmitting them to the Ethernet output interface unit. And a frame multiplexing unit for, converting a preset Ethernet frame data from the video output device connected to the image input interface, characterized in that that can be transmitted. According to the above configuration, the image input interface unit receives the data from the external image input device such as the video camera through the image input interface, and further sends the data to the frame generation unit. The frame generation unit holds two types of header information to be preset, one for video data and the other for audio data. The frame generation unit adds the header information for the video data to the beginning of the data received from the video input interface. Then, a frame that can be output via Ethernet is constructed and passed to the frame multiplexing unit. The frame multiplexing unit multiplexes the frame generated by the frame generating unit and the data requested to be transmitted by the Ethernet device driver on the host computer through the computer interface unit, and passes it to the Ethernet output interface unit. The Ethernet output interface unit sequentially sends the frames received from the frame multiplexing unit to the Ethernet. From the above, among the frames received from the Ethernet input interface, for frames with a frame pattern set for specific video and audio, the frame header on the adapter device does not go through the software on the host computer. Can be removed and raw video data or audio data can be taken out and passed to the video output interface section or the audio output interface section, respectively, and the delay from frame reception to video output or audio output can be reduced. . Also, video data from a video input device connected to the video input interface section can be directly framed and sent out on the adapter device without going through software on the host computer. It is possible to reduce the delay until frame transmission. Furthermore, the audio data from the audio input device connected to the audio input interface unit can be directly framed and sent out on the adapter device without going through the software on the host computer. It is possible to reduce the delay until frame transmission.

Further, in the present invention as set forth in claim 5, the frame filter section includes a filter section for performing pattern matching of frames, a pattern memory section for holding pattern information, and a frame having a matched pattern. The delete offset memory unit includes a header deleting unit having a function of deleting the header information from the beginning by the amount held by the offset memory unit, and a memory updating unit for updating the memory information of the pattern memory unit and the memory information of the delete offset memory unit. The device driver for Ethernet on the host computer can update the data in the pattern memory unit by using the memory updating unit through the computer interface unit, and can similarly update the data in the deleted offset memory unit. Characterized by what is possible That. According to the above configuration, the application on the host computer specifies the frame pattern information to be filtered and the length of the header to be deleted for the frame that matches the pattern, through the Ethernet device driver. From the above,
You can specify the pattern of the protocol header that performs pattern matching, and you can also specify the offset value that should be deleted even for the header length that differs depending on the protocol used, so audio and video data can be encapsulated by various protocols. Even if the information is converted and transmitted, it can be dealt with by appropriately setting the pattern information and the deletion offset information. Since the pattern of the protocol header for pattern matching can be specified according to the protocol used at that time, and the offset value to be deleted can be specified even for the header length that differs depending on the protocol used, Even when audio data and video data are encapsulated and transmitted by various protocols, it is possible to deal with them by appropriately setting the pattern information and the deletion offset information.

According to the present invention of claim 6, the frame generation section includes a header information memory section for holding header information to be set at the head of the received data, and a header information memory section at the head of the received data. A device having a function of adding data and a memory updating unit for updating the data in the header information memory unit are provided, and the Ethernet device driver on the host computer uses the memory updating unit through the computer interface unit. The data in the header information memory unit can be updated. According to the above configuration, the application on the host computer specifies the header information of the frame to be generated through the Ethernet device driver. As described above, it is possible to specify the information of the protocol header when performing frame generation, so it is possible to handle it by appropriately setting the header information even when transmitting audio data and video data encapsulated by various protocols. Become.

Further, according to the invention of claim 7, the voice frame extracted by the frame filter unit is received among the Ethernet frames received through the Ethernet input interface unit, and is preset for the received frame. A protocol process having a function of transmitting to the voice output interface unit after performing the protocol process, and a function of transmitting to the frame generation unit after performing a predetermined protocol process on the data received in the voice input interface unit It is possible to perform protocol processing on the frame extracted by the frame filter section and directly transmitted to the audio output interface section, and also on the frame received from the audio input interface section. What you can do And it features. According to the above configuration, only the unnecessary header part is deleted in the frame filter part, and the header part of the protocol that requires protocol processing such as sequence number processing is not deleted in the frame filter part but is deleted in the protocol processing part. Perform processing. The protocol processing section deletes the header section after performing necessary processing according to the protocol, extracts the raw audio data, and transmits it to the audio output interface section. Also, the protocol processing unit receives the data input from the voice input interface unit, performs protocol processing according to the protocol such as header addition, and passes the data to the frame generation unit. Since the protocol processing unit is the same as the protocol processing unit that is executed when receiving from the Ethernet input interface, for example, a header obtained by adding 1 to the sequence number set in the frame from the Ethernet input interface is input as voice. Transmission and reception such as giving to the data from the interface unit and passing it to the frame generation unit are related to each other, and it is possible to dynamically change the header. Also, when it is necessary to return a response to a frame received from the Ethernet input interface unit, a frame for that protocol is created inside the protocol processing unit, and the header of the lower protocol is added in the frame generation unit. By doing so, it is possible to perform preset protocol processing without going through the protocol processing unit on the host computer.

According to the invention of claim 8, in addition to the protocol processing unit for audio data, a video frame extracted by a frame filter unit is received from an Ethernet frame received through an Ethernet interface unit, After performing a preset protocol process on the received frame and sending it to the video output interface unit, and after performing a predetermined protocol process on the data received in the video input interface unit, a frame generation unit It is equipped with a video protocol processing unit that has the function of transmitting to the frame, and it is possible to perform protocol processing on the frame extracted by the frame filter unit and directly transmitted to the video output interface unit. Also for received frames Characterized that it is possible to perform Tokoru process. According to the above configuration, only the unnecessary header part is deleted in the frame filter part, and the header part of the protocol that requires protocol processing such as sequence number processing is not deleted in the frame filter part but is deleted in the protocol processing part. Perform processing. The protocol processing section deletes the header section after performing necessary processing according to the protocol, extracts the raw audio data, and transmits it to the audio output interface section. Also, the protocol processing unit receives the data input from the voice input interface unit, performs protocol processing according to the protocol such as header addition, and passes the data to the frame generation unit. From the above, transmission and reception are related to each other by the protocol processing unit not only for audio but also for video,
It is also possible to dynamically change the header. Moreover, not only for audio but also for video, when a response needs to be returned for a frame received from the Ethernet input interface unit, a frame for that protocol is created inside the protocol processing unit, and By adding a lower-layer protocol header in the frame generation unit, it is possible to perform preset protocol processing without going through the protocol processing unit on the host computer.

According to a ninth aspect of the invention, the pattern information set in the frame filter section is TCP /
It is characterized in that it is only the own port number information in IP communication or UDP / IP communication. According to the above configuration, when the audio data and the video data are transmitted / received through the TCP / IP communication or the UDP / IP communication, the pattern information set in the frame filter unit is the TCP / IP communication or the UDP / IP communication.
In IP communication, only the own port number set in the TCP header or UDP header is used. Frames are filtered only by the own port number without referring to other IP address information. As described above, the audio data and the video data can be filtered only by the pattern matching having a very small size of the own port number (for example, 2 bytes).

According to the tenth aspect of the invention, the information set in the frame filter section is H.264. 323
Is a port number dynamically assigned by the call control by the Internet telephone such as H.264. A transmission MAC address, a reception MAC address, a transmission IP address, a transmission port number, a reception IP address, and its own port number, which are dynamically assigned by call control by an Internet telephone such as 323. According to the above configuration, like the Internet telephone, the call control is performed with the other computer, the port to be used, the other IP
In a protocol in which the address is determined as a result of performing call control, after the call setup is completed, the transmission MAC address, the reception MAC address, the transmission IP address, the transmission port number, the reception IP address, and the own port number are set in the frame. It can be set in the header information memory of the generation unit, and the own port number can be set in the pattern memory unit in the frame filter unit.

[0024]

(First Embodiment) FIG. 1 is a block diagram showing a configuration of a first embodiment of the present invention. In FIG. 1, reference numeral 1 is an Internet telephone service application. Reference numeral 2 denotes a protocol processing unit on the host computer, which is software for processing a protocol such as TCP / IP. Reference numeral 3 denotes a device driver for the network adapter device of the present invention, and the protocol processing unit 2 on the host computer can transmit and receive data via the device driver 3.

Reference numeral 4 is a network adapter device of the present invention, which is installed in an expansion slot of the host computer and plays a role of connecting the host computer to the Ethernet.

Reference numeral 5 denotes a computer interface unit, which includes a network adapter device 4 and a device driver 3.
And the device driver 3 provides an interface for accessing the network adapter device 4.
For example, a PCI bus may be mentioned as a suitable example.

A frame filter unit 6 receives a frame of the received data 11 input to the Ethernet input interface unit 9, collates the preset pattern information with the frame, and if they match, from the beginning of the frame. The data 15 deleted by a preset amount is sent to the voice output interface unit 7. If they do not match, the data 13 is passed through the computer interface unit 5 to the device driver 3 without being deleted from the beginning of the frame.

When the protocol processing unit 2 transmits data, the data is passed to the device driver 3, and the device driver 3 transfers the data 14 via the computer interface unit 5 to the Ethernet output interface unit 8
Pass to. The Ethernet output interface unit 8 transmits the transmission data 10 to Ethernet.

As described above, according to the above embodiment, unlike the conventional apparatus (shown in FIG. 15), the contents of the frame matching the preset pattern are processed by the device driver 3 on the host computer or the protocol processing. It is possible to directly deliver to the voice output interface unit 7 on the adapter without passing through the unit 2 and the Internet telephone service application 1.

FIG. 2 is a diagram showing the contents of data processed by the above configuration. In the figure, UDP / IP communication is used as an example of whether the frame received by the Ethernet input interface unit 9 is pattern-matched by the frame filter unit 6 and is deleted by a specified number of bytes from the beginning of the frame. The case is shown.

As shown in FIG. 2, the frame of the reception data 11 input to the Ethernet input interface unit 9 includes a MAC header, a UDP header, an IP header and actual data (voice data, for example). The Ethernet input interface unit 9 receives the received data 11 from the MA
Remove the C header and pass the data to the frame filter unit 6. The frame filter unit 6 checks the matching with a preset specific pattern. For example, as an example of the specific pattern, 4 bytes are set to 0x10100101 from an offset of 20 bytes from the beginning. In this example, the IP header is 20 bytes, so 4 bytes from the offset 20 bytes are UDP
Frames that are the transmission port number and the own port number of the header and are 0x1010 and 0x0101 are extracted by the frame filter unit 6, and other frames are passed to the device driver 3 via the computer interface unit 5. .

For frames that match the pattern,
Further, only the data 15 after the portion indicated by the preset deletion offset value is passed to the audio output interface unit 7. For example, if 28 bytes are set as the deletion offset value, the IP header is 20 bytes and the UDP header is 8 bytes. Therefore, pass the raw data with the IP header and UDP header removed to the audio output interface unit 7. Is possible.

As described above, according to the first embodiment of the present invention, only the frame having the preset specific pattern is extracted from the frames of the reception data 11 input to the Ethernet input interface unit 9. It is possible to remove the protocol header at the beginning of the frame, and pass only the raw data 15 to the audio output interface unit 7. In particular, when the raw data 15 is voice data, unlike the Internet telephone or the like, it is possible to directly pass the data to the voice output interface unit 7 on the adapter device 4 without going through software on the computer. Because there is
It is possible to reduce the delay from the frame reception to the voice output, so that it becomes possible to provide a high quality voice reception service.

(Second Embodiment) FIG. 3 is a block diagram showing a second embodiment of the present invention. In each of the second and subsequent embodiments, the same components as those in the first embodiment (shown in FIG. 1) are designated by the same reference numerals, and detailed description thereof will be omitted. In FIG. 3, reference numeral 22 is a video output interface unit. In the present embodiment, the frame filter unit 6 includes two types of pattern information, that is, the pattern information for video and the pattern information for audio, and two types of the deletion offset value for video and the deletion offset value for audio. The frame of the received data 11 from the Ethernet input interface unit 9 is received based on the deletion offset value, and if the pattern information for the video and the frame match, the data is deleted from the beginning of the frame based on the deletion offset value for the video. After that, the data is passed to the video output interface unit 22.

If the frame from the Ethernet input interface unit 9 matches the voice pattern information, data is deleted from the beginning of the frame based on the voice deletion offset value, and then the data is passed to the voice output interface unit 7. . If the frame from the Ethernet input interface unit 9 does not match the video pattern information or the audio pattern information, the head of the frame is not deleted and the device driver on the host computer is directly passed through the computer interface unit 5. Pass the data to 3 (see FIG. 1).

As described above, according to the second embodiment, only the frame having the preset specific pattern is received by the Ethernet input interface unit 9 as the received data 1
It is possible to take out from one frame, and further, it is possible to take out the protocol header at the beginning of the frame and take out raw data.

Further, by holding the pattern information and the deletion offset value for the video and audio, when the pattern information for the video matches, the protocol header at the head of the frame is set based on the deletion offset value for the video. It is possible to remove and pass only the raw video data to the video output interface unit 22. Therefore, since it is possible to directly pass the video data to the video output interface unit 7 on the adapter device 4 without passing through software on a computer like a general video distribution application, from frame reception to video output. It is possible to reduce the delay of, and thus it is possible to provide a high quality video reception service.

Similarly, for voice, when the voice pattern information matches, the protocol header at the beginning of the frame is removed based on the voice deletion offset value, and only the raw voice data is output to the voice output interface unit. It is possible to pass to 7. Therefore, since audio data can be directly passed to the audio output interface section on the adapter device 4 without passing through software on a computer like a general audio distribution application, from frame reception to audio output. It is possible to reduce the delay and thus provide a high quality voice receiving service.

In addition, particularly, since a large amount of video data is transferred to the video output interface section via software on a computer as in the conventional method, CPU resources, memory resources, etc. on the computer are wasted. According to the method of the present invention, since the video data is directly passed to the video output interface unit 22 on the adapter device 4, it is possible to receive the video data without wasting CPU resources, memory resources, etc. on the computer. Become.

(Third Embodiment) FIG. 4 shows a third embodiment of the present invention. In FIG. 4, the Ethernet input interface unit 9, the frame filter unit 6, and the audio output interface unit 7 have the same configurations as those described in the first embodiment.
It has a function of extracting only the audio frame from the frame based on preset pattern information and transmitting it to the audio output interface unit 7.

In FIG. 4, reference numeral 23 denotes a voice input interface unit, which can be connected to an external voice input device such as a microphone, and passes data input from the external voice input device to the frame generation unit 24.

The frame generation unit 24 holds preset header information, adds the header information to the head of the data from the voice input interface unit 23, forms an Ethernet frame, and passes it to the frame multiplexing unit 25. .

The frame multiplexing unit 25 is provided with the Ethernet frame passed from the computer interface unit 5,
The Ethernet frame passed from the frame generation unit 24 is multiplexed and passed to the Ethernet output interface unit 8.

As described above, according to the third embodiment, only the frame having the preset specific pattern can be extracted from the 11 frames of the received data input to the Ethernet input interface unit 9. Yes, it is possible to remove the protocol header at the beginning of the frame and pass only the raw data to the audio output interface unit 7. In particular, when the raw data is voice data, it is possible to directly pass the data to the voice output interface unit 7 on the adapter device 4 without going through software on the computer, unlike the Internet telephone. Therefore, it is possible to reduce the delay from the frame reception to the voice output, and thus it is possible to provide a high quality voice reception service.

Furthermore, by adding preset header information to the data received by the voice input interface section 23, it is possible to generate an Ethernet frame by the frame generation section 24, and further by the frame multiplexing section 25, Ethernet. The output interface unit 8 can send the frame to the Ethernet. Since the above processing is directly processed on the adapter device 4 without using software on the computer, a delay from voice input to Ethernet transmission is delayed as compared with the conventional method of transmitting to Ethernet via software on the computer. Can be reduced.

(Fourth Embodiment) FIG. 5 shows a fourth embodiment of the present invention. In FIG. 5, the frame of the received data input to the Ethernet input interface unit 9 is passed to the frame filter unit 6. The frame filter unit 6 holds two types of pattern information for video and audio in advance. When the pattern information matches the pattern information for video based on these pattern information, a preset deletion offset value for video is set. After deleting only the data from the beginning of the frame, the remaining data is passed to the video output interface unit 22. When it matches the voice pattern information, the data corresponding to the preset voice deletion offset value is deleted from the beginning of the frame, and the remaining data is passed to the voice output interface unit 7.

In FIG. 5, the voice input interface unit 23 has the same configuration as that described in the third embodiment, and an external voice input device such as a microphone can be connected to the voice input interface unit 23. The generated data is passed to the frame generator 24.

Reference numeral 26 denotes a video input interface unit, which can be connected to an external video input device such as a video camera, and passes data input from the external video input device to the frame generation unit 24.

The frame generator 24 holds two types of header information, which are preset video header information and audio header information, and when data is received from the audio input interface unit 7, it is added to the beginning of the data. , Adds header information for voice, forms an Ethernet frame, and passes it to the frame multiplexing unit 25. When data is received from the video input interface unit 26, video header information is added to the beginning of the data to form an Ethernet frame,
It is passed to the frame multiplexing unit 25.

The frame multiplexing unit 25 is provided with the Ethernet frame passed from the computer interface unit 5,
The Ethernet frame passed from the frame generation unit 24 is multiplexed and passed to the Ethernet output interface unit 8.

As described above, according to the fourth embodiment, only the frame having the preset specific pattern can be extracted from the frames of the received data 11 input to the Ethernet input interface section 9. It is possible to remove the protocol header at the beginning of the frame and pass only the raw data to the video output interface unit 22 or the audio output interface unit 7. Therefore, it is possible to directly pass the data to the video output interface unit 22 or the audio output interface unit 7 on the adapter device 4 without passing through the software on the computer, so that from the frame reception to the video output or the audio output. It is possible to reduce the delay and provide a high-quality video reception service and audio reception service.

Furthermore, by adding preset header information to the data received by the video input interface unit 26 or the audio input interface unit, the frame generation unit 24 can generate an Ethernet frame, and further the frame can be generated. The multiplexing unit 25 and the Ethernet output interface unit 8 can send the frame to Ethernet. The above process does not require software on the computer
Since the processing is performed directly on the adapter device 4, it is possible to reduce the delay from the video input or the audio input to the Ethernet transmission as compared with the conventional method of transmitting to the Ethernet via the software on the computer.

Since the amount of video data is large, if the video data is processed by software on a computer, CPU resources and memory resources on the computer are wasted. However, according to the fourth embodiment, It is possible to use video and audio services without using resources.

(Embodiment 5) FIG. 6 shows a fifth embodiment of the present invention. FIG. 6 shows the internal structure of the frame filter unit 6, which is composed of the following units.

Reference numeral 27 denotes a memory updating unit, and the device driver 3 (shown in FIG. 1) on the computer uses the memory updating unit 27 via the computer interface unit 5 to determine the contents of the deletion offset memory unit 28 and It is possible to update the contents of the pattern memory unit 29.

The contents set in the pattern memory unit 29 are referred to by the filter unit 31 and used when the Ethernet input interface unit 9 receives the received data to determine whether or not they match the pattern.

The contents held in the pattern memory unit 29 include an offset value for checking pattern matching,
That is, it includes a value that holds the number of bytes from the beginning of the frame to be checked, and a bit pattern that is actually checked.

The filter unit 31 refers to the contents of the pattern memory unit 29 and determines whether or not the frame received from the Ethernet input interface unit 9 has the same bit pattern from the offset value. If they match, pass the frame to the header deletion section,
If they do not match, the frame is directly transmitted to the device driver 3 via the computer interface unit 5.
Pass to.

The header deletion unit 30 refers to the contents held in the deletion offset memory unit 28 and refers to the filter unit 3
For the frame received from 1, after deleting only the value of the deletion offset memory unit 28 from the beginning of the frame, the remaining data is passed to the audio output interface unit 7.

As described above, according to the fifth embodiment, the Ethernet device driver 3 on the host computer uses the memory updating means 27 through the computer interface section 5 so that the data of the pattern memory section 29 can be obtained. Can be updated, and similarly, the data in the deletion offset memory unit 28 can be updated. Therefore, the pattern of the protocol header for pattern matching can be specified according to the protocol used at that time. Also, since the offset value to be deleted can be specified even for the header length that differs depending on the protocol used, even if audio data and video data are encapsulated and transmitted by various protocols, pattern information and deletion By setting the offset information appropriately, It can become.

(Sixth Embodiment) FIG. 7 shows a sixth embodiment of the present invention. FIG. 7 shows the internal structure of the frame filter unit 24.

Reference numeral 32 denotes a memory updating unit, and the device driver 3 (shown in FIG. 1) on the computer via the computer interface unit 5 updates the contents of the header information memory unit 33 using the memory updating unit 32. It is possible.

The contents set in the header information memory unit 33 are referred to by the generation unit 34, and when the data from the voice input interface unit 7 is received, the contents of the header information memory unit 33 are added to the beginning of the data. To be done.

The contents held in the header information memory unit 33 include the header of the Ethernet frame, that is, MA.
The C header part and the header part of the protocol used at that time. For example, when performing TCP / IP communication, a MAC header containing its own MAC address and destination MAC address, a TCP header containing its own IP address and destination IP address, and a lower protocol type (TCP in this example), its own port It is each protocol header of the TCP header including the number and the destination port number.

The generation unit 34 adds the contents set in the header information memory unit 33 to the beginning of the data received from the voice input interface unit 7 as it is to form an Ethernet frame. This frame is passed to the frame multiplexing unit 25.

In the frame multiplexing unit 25, the Ethernet frame which the device driver 3 on the computer has requested to transmit via the computer interface unit 5,
Frame multiplexing is performed so that the Ethernet frames requested to be transmitted by the frame generation unit 24 do not mix, and the frames are passed to the Ethernet output interface unit 8. The Ethernet output interface unit 8 sends the passed Ethernet frame to the Ethernet.

As described above, according to the sixth embodiment, the device driver 3 for Ethernet on the host computer uses the memory updating means 32 through the computer interface section 5 to make the header information memory section 33. Data can be updated, it is possible to specify the protocol header for adding headers according to the protocol used at that time. Therefore, it is possible to encapsulate and transmit audio data and video data using various protocols. Is possible.

(Seventh Embodiment) FIG. 8 shows a seventh embodiment of the present invention. As shown in FIG. 8, the seventh embodiment has a configuration in which a protocol processing unit 35 is provided in addition to the components of the third embodiment (described in FIG. 4) described above, and the protocol processing unit 35 is a frame filter unit. Performs protocol processing on the data received from. Conversely, protocol processing is performed on the data received from the voice input interface unit 23.

FIG. 9 is an explanatory view of the processing in the seventh embodiment. Ethernet input interface section 8
The RTP / UDP / IP communication is an example of how the frame received in step 3 is subjected to pattern matching processing in the frame filter section 6, further a specified number of bytes are deleted from the beginning of the frame, and passed to the protocol processing section 35. It is shown when it is used.

As shown in FIG. 9, the frame of the received data input to the Ethernet input interface section 9 includes a MAC header, a UDP header, an IP header, an RTP header and actual data (eg voice data). There is. The Ethernet input interface unit 9 removes the MAC header and passes the data to the frame filter unit 6.
The frame filter unit 6 checks the matching with a preset specific pattern. For example, as an example of the specific pattern, 4 bytes are set to 0x10100101 from an offset of 20 bytes from the beginning. In this example, the IP header is 20 bytes, so 4 bytes from the offset 20 bytes are the transmission port number and the own port number of the UDP header, and
Frames such as 0x1010 and 0x0101 are extracted by the frame filter unit 6, and other frames are passed to the device driver 3 via the computer interface unit 5.

For frames that match the pattern,
Further, only the data after the portion indicated by the preset deletion offset value is passed to the protocol processing unit 35. For example, if 28 bytes is set as the deletion offset value, the IP header is 20 bytes and the UDP header is 8 bytes, so the rest of the IP header and UDP header (RTP header and voice data) is processed by protocol. Hand it over to department 35.

In the protocol processing unit 35, processing based on a preset protocol is performed. This 7th
In the embodiment, the RTP header of the data received from the frame filter unit 6 is referred to, the processing based on the RTP is performed,
After deleting the RTP header, raw audio data can be passed to the audio output interface unit 7.

Further, as shown in FIG. 10, the protocol processing section 35 receives the data input from the voice input interface section 23, performs protocol processing according to the protocol such as header addition, and converts the data into the frame generation section 2
Pass to 4. Since the protocol processing unit 35 is the same as the protocol processing performed in the reception from the Ethernet input interface unit 9, for example, 1 is added to the sequence number set in the frame from the Ethernet input interface unit 9. Transmission and reception such as adding a header to the data from the voice input interface unit 7 and passing the data to the frame generation unit 24 are related to each other, and it is possible to dynamically change the header.

Also, when it is necessary to return a response to a frame received from the Ethernet input interface unit 9, a frame of the protocol is created inside the protocol processing unit 35, and the frame generation unit 24 further By adding the header of the lower protocol, the protocol processing unit 3 on the host computer
It is possible to perform preset protocol processing without going through 5.

(Embodiment 8) FIG. 11 shows an eighth embodiment of the present invention.
FIG. As described in the seventh embodiment, since the voice protocol processing unit 35 remains the same, the voice data input from the Ethernet input interface unit 9 is extracted by the frame filter unit 6 and the voice protocol is used. After the protocol processing is performed by the processing unit 35, it is passed to the audio output interface unit 7.

On the contrary, the data input from the voice input interface unit 23 is protocol-processed by the protocol processing unit 35, and then converted into an Ethernet frame by the frame generation unit 24.
It is sent to the Ethernet output interface unit 8 via 5.

In FIG. 11, reference numeral 36 is a video protocol processing unit, which performs protocol processing on the data received from the frame filter unit 6. Conversely, protocol processing is performed on the data received from the video input interface unit 26.

The frame of the data received by the Ethernet input interface unit 9 includes various protocol headers (UDP header, IP header, RTP header, etc.) and actual data (for example, voice data) in addition to the MAC header. ing. The Ethernet input interface unit 9 removes the MAC header and passes the data to the frame filter unit 6.
The frame filter unit 6 checks the matching between a preset specific pattern for video and a specific pattern for audio. Frames that do not match either pattern are passed to the device driver 3 via the computer interface unit 5. For the frame that matches the video pattern, only the data after the portion indicated by the preset video deletion offset value is passed to the video protocol processing unit 36.

In the video protocol processing section 36,
Processing based on a preset protocol is performed. The video protocol processing unit 36 can pass the raw video data to the video output interface unit 22 after removing the protocol header.

Further, in the frame filter section 6, the frame which matches the specific pattern for voice is processed in the same manner as the case shown in the seventh embodiment.

Further, the video protocol processing unit 36 receives the data input from the video input interface unit 26, performs protocol processing according to the protocol such as header addition, and passes the data to the frame generation unit 24.

Since the video protocol processing unit 36 is the same as the video protocol processing performed in the reception from the Ethernet input interface unit 9, for example, the sequence number set in the frame from the Ethernet input interface unit 9 is used. On the other hand, the header added with 1 is added to the data from the video input interface unit 26, and the data is sent to the frame generation unit 24 so that the transmission and reception are related to each other, and the processing of dynamically changing the header is also possible.

Also, when it is necessary to return a response to the frame of the received data of the Ethernet input interface unit 9, the frame of the protocol is created inside the video protocol processing unit 36, and the frame is further generated. By adding the header of the lower protocol in the unit 24, it is possible to perform preset protocol processing without going through the protocol processing unit 2 on the host computer.

(Ninth Embodiment) FIG. 12 shows a ninth embodiment of the present invention.
FIG. In FIG. 12, 37 indicates the own port number field.

As shown in FIG. 12, the own port number field 37 is in the same place in the case of TCP / IP communication and the case of UDP / IP communication, and its size is 2 bytes.

In the method of holding the data of only 2 bytes in the frame filter unit 6 and judging the pattern,
Since only a very small amount (2 bytes) of the fixed offset position needs to be matched, the pattern determination can be performed very quickly, and the pattern determination circuit can be easily implemented at low cost.

Further, one connection in TCP / IP communication and UDP / IP communication can be identified by the protocol, own IP address, partner IP address, own port number, partner port number, and if it arrives on its own computer. The Ethernet frame can be identified by the protocol and its own port number. Since it is rare to use the same port number for different protocols, it is possible to retrieve only the desired data even if frame filtering is executed using only the own port number.

(Embodiment 10) FIG. 13 is a diagram showing a tenth embodiment of the present invention. In FIG.
The Internet telephone protocol typified by H.323 is call control protocol software 4 on a computer.
0 is used to set the connection used for the call with the other party of the call. In Internet telephony, this connection typically uses a TCP / IP connection or a UDP / IP connection.

In TCP / IP or UDP / IP, a connection is designated by a destination IP address, destination port number, own IP address, and own port number. Therefore, setting a connection by the call control protocol means that the port number used with this communication partner is determined.

According to the tenth embodiment, the frame filter section 6 sets the own port number of the connection for such a dynamically set connection.

When such a connection is established,
Further, it becomes possible to configure the MAC header, the IP header, and the UDP header including the destination MAC address as shown in FIG. 14, and the internet telephone application, through the device driver 3, transmits the information of these MAC header, IP header and UDP header. Can be set in the frame generation unit 24 via the computer interface unit 5.

Also, in many Internet telephone applications, it is not just UDP / IP communication but RT
P / UDP / IP communication is used. Since the RTP requires processing such as sequence number allocation, it cannot be handled only by the header information that is semi-fixedly set in the frame generation unit 24, but the protocol processing unit 35 provided on the adapter device 4 is used as the RTP. In this case, the frame filter unit 6 performs filtering based on its own port number, deletes the UDP header and IP header, and the protocol processing unit 35 interprets, processes, and deletes the RTP header. , To the audio output interface unit 7. The processing of the protocol stack used in Internet telephony applications can thus be divided up and processed.

[0093]

According to the present invention, since the data is directly transferred to the input / output interface section on the adapter, it is possible to reduce the delay from the frame reception to the output of audio and video, and therefore, It is possible to obtain the effect that it is possible to provide a high quality audio and video reception service.

In particular, although there is a large amount of video data, since the video data is passed directly to the video output interface section on the adapter, it is possible to receive the video data without wasting CPU resources, memory resources, etc. on the computer. Becomes

[Brief description of drawings]

FIG. 1 is a system configuration diagram of an Internet telephone service in a first embodiment of a network adapter device of the present invention.

FIG. 2 is an explanatory diagram of a frame filter unit according to the first embodiment of the present invention.

FIG. 3 is a configuration diagram of a network adapter device according to a second embodiment of the present invention.

FIG. 4 is a configuration diagram of a network adapter device according to a third embodiment of the present invention.

FIG. 5 is a configuration diagram of a network adapter device according to a fourth embodiment of the present invention.

FIG. 6 is a configuration diagram of a frame filter unit in a network adapter device according to a fifth embodiment of the present invention.

FIG. 7 is a configuration diagram of a frame generation unit in a network adapter device according to a sixth embodiment of the present invention.

FIG. 8 is a configuration diagram of a network adapter device according to a seventh embodiment of the present invention.

FIG. 9 is an explanatory diagram of processing of a frame filter unit and a protocol processing unit according to the seventh embodiment of the present invention.

FIG. 10 is an explanatory diagram of processing of a frame generation unit according to the seventh embodiment of the present invention.

FIG. 11 is a configuration diagram of a network adapter device according to an eighth embodiment of the present invention.

FIG. 12 is an explanatory diagram of the operation of the frame filter unit according to the ninth embodiment of the present invention.

FIG. 13 is a configuration diagram of an operation of the network adapter device according to the tenth embodiment of the present invention.

FIG. 14 is an explanatory diagram of processing of a frame generation unit and a protocol processing unit according to the tenth embodiment of the present invention.

FIG. 15 is a system configuration diagram of an Internet telephone service using a conventional network adapter device.

[Explanation of sign]

1 Internet telephone service application 2 Protocol processing unit on the host computer 3 Device driver 4 Network adapter device 5 Computer interface section 6 Frame filter section 7 Audio output interface section 8 Ethernet output interface section 9 Ethernet input interface section 10 Transmission data to Ethernet 11 Received data from Ethernet 12 Data sent to audio output device 13 Data that did not match the filter pattern 14 Transmission data from host computer 15 Data that matches the filter pattern 16 Ethernet device driver 17 Audio device driver 18 Ethernet adapter device 19 Audio adapter device 20 Ethernet interface section 21 Voice interface section 22 Video output interface section 23 Voice input interface section 24 frame generator 25 frame multiplexer 26 Video input interface section 27 Memory updating means of frame filter section 28 Deleted offset memory section 29 Pattern memory section 30 Header deletion part 31 Filter section 32 memory update means of frame generation unit 33 Header information memory section 34 Generator 35 Protocol processing unit 36 Video Protocol Processor

Continuation of front page (58) Fields surveyed (Int.Cl. ⁷ , DB name) H04L 12/40 H04L 12/44

Claims (10)

(57) [Claims] 1. A network adapter device for connecting a computer and an Ethernet, for connecting a computer to transmit and receive various control information and data, and for connecting to the Ethernet to receive an Ethernet frame. Ethernet input interface section, an Ethernet output interface section for transmitting an Ethernet frame, an audio output interface section for connecting to an audio output device and outputting audio data, and a preset pattern for a frame input from Ethernet. If the patterns match, the data corresponding to the preset deletion offset value is removed from the beginning of the frame and the remaining part is sent to the audio output interface section. The frame filter section has the function of transmitting the frame to the computer interface section as it is. Only the input frame from the Ethernet having a preset pattern is extracted in the frame filter section, and only the frame is used as the audio output interface. A network characterized by outputting and passing other frames to the protocol software on the computer through the computer interface unit, and for the output frame to the Ethernet, the data passed from the computer interface unit is output as it is through the Ethernet output interface unit. Adapter device. 2. In addition to the Ethernet input interface unit, the Ethernet output interface unit, the computer interface unit, the audio output interface unit, a video output interface unit for connecting to a video output device and outputting video data. Then, the difference between the preset audio pattern and the video pattern for the input frame from the Ethernet input interface section is determined, and if they match the audio pattern, the input frame is set to the audio output interface section. It is equipped with a frame filter unit that has the function of transmitting data and transmitting an input frame to the video output interface unit if it matches the image pattern, and transmitting data to the computer interface unit if it does not match either pattern. from Of the input frames, only those with a preset pattern are extracted by the frame filter section, data corresponding to the preset offset value is removed from the beginning of the frame, and the remaining portion is output to the audio output interface section or video output interface section. However, the other frames are directly passed to the protocol software on the computer through the computer interface unit, and for the output frame to the Ethernet, the data passed from the computer interface unit is directly output through the Ethernet output interface unit. The network adapter device according to item 1. 3. The Ethernet input interface unit, the frame filter unit, and the audio output interface unit have a filtering function based on preset pattern information with respect to a frame received from Ethernet, and input from an audio input device. A voice input interface unit having a function of receiving data and transmitting it to the frame generation unit; and a frame generation unit having a function of adding preset header information to the data received from the voice input interface and transmitting it to the frame multiplexing unit. Connected to the audio input interface, equipped with a frame multiplexer that receives the transmission frame from the computer interface and the transmission frame from the frame generation unit, multiplexes them, and transmits them to the Ethernet output interface unit The network adapter device according to claim 1, wherein the data from the audio output device that has been converted can be converted into a preset Ethernet frame and transmitted. 4. The Ethernet input interface unit, the frame filter unit, the audio output interface unit, and the video output interface unit have a filtering function based on preset pattern information for a frame received from Ethernet, A video input interface section that has the function of receiving input data from the video input device and transmitting it to the frame generation section, and a function that adds preset header information to the data received from the video input interface and sends it to the frame multiplexing section And a frame multiplexing unit having a function of receiving a transmission frame from the computer interface and a transmission frame from the frame generation unit, multiplexing the received frames, and transmitting the multiplexed frames to the Ethernet output interface unit. Network adapter device according to claim 2, characterized in that that can be converted into a preset Ethernet frame data from the video output device connected to the image input interface transmits. 5. The frame filter section is held by a deletion offset memory section for a pattern section for holding pattern information, a filter section for executing pattern matching of frames, and a pattern memory section for holding pattern information. An Ethernet device driver on the host computer, which includes a header deleting unit having a function of deleting header information from the beginning by a minute, and a memory updating unit for updating memory information of the pattern memory unit and memory information of the deleted offset memory unit. The data can be updated in the pattern memory unit by using the memory updating unit through the computer interface unit, and the data in the deletion offset memory unit can be updated in the same manner. The net according to any one of 1 to 4. Network adapter device. 6. The frame generation unit has a header information memory unit for holding header information to be set at the beginning of received data, and a generation unit having a function of adding data of the header information memory unit to the beginning of received data. And a memory updating unit for updating the data in the header information memory unit. The device driver for Ethernet on the host computer updates the data in the header information memory unit by using the memory updating unit through the computer interface unit. claim 3, characterized that it is possible to
Alternatively, the network adapter device according to item 4 . 7. An audio frame extracted by a frame filter unit is received from an Ethernet frame received through an Ethernet input interface unit,
After the preset protocol processing is performed on the received frame, the function to send to the audio output interface section, and after performing the predetermined protocol processing on the data received in the audio input interface section, the frame A protocol processing unit having a function of transmitting to the generation unit is provided, and it is possible to perform protocol processing on the frame extracted by the frame filter unit and directly transmitted to the audio output interface unit. The network adapter device according to any one of claims 1 to 6, wherein protocol processing can be performed on the received frame. 8. In addition to the audio data protocol processing unit, an Ethernet frame received through an Ethernet interface unit receives a video frame extracted by a frame filter unit, and is preset for the received frame. The video protocol having a function of transmitting the data to the video output interface unit after performing the protocol processing, and performing a predetermined protocol process on the data received in the video input interface unit, and then transmitting the data to the frame generation unit. It is equipped with a processing unit and can perform protocol processing on the frames extracted by the frame filter unit and directly sent to the video output interface unit, and also on the frames received from the video input interface unit. Can do Claim 2,4～, characterized in that
6. The network adapter device according to any one of 6. 9. The network according to claim 1, wherein the pattern information set in the frame filter unit is only own port number information in TCP / IP communication or UDP / IP communication. Adapter device. 10. The information set in the frame filter unit is H.264. It is a self-port number dynamically assigned by the call control by the Internet telephone such as H.323, and the information set in the frame generation unit is H.323.
Sending MAC address and receiving M dynamically assigned by call control by Internet telephone such as H.323.
The network adapter device according to any one of claims 7 to 9, wherein the network adapter device is an AC address, a transmission IP address, a transmission port number, a reception IP address, and its own port number.